Lessons of the Week
This week was a bit slow. I was fasting on a diet for 5 days, which reduced my brain power a bit, but I dropped 8 lbs and feel awesome. My wife lost 3lbs just by watching me not eat. The value of positive externalities.
Next week is World Agritech. I am there only to find leaders in regenerative ag, nutrient density, and metabolic health. We are building a vertical network of solutions to define the "Tech Stack" of food is health. We are moving to define the key companies from dirt to healthcare that comprise the food is health solution.
Consumers, not CPGs, control diet, and metabolic health
Is grocery in the business of selling shelf space or nutrition
CPGs do not know the economic value of nutrient density
CPG product development is 90% value engineering, not nutrition quality
NPK undermines biologics in ways we do not understand
Soil health is a 5 year strategy, not seasonal
Incumbent powers fail to realize no one has access to a better LLM than what billions of people around the world get for $0-$20/month.
If regenerative agriculture = nutrient density and
nutrient density = health, then regenerative agriculture = health (h/t Raviv Turner)Anti-obesity medications (AOM) cost more than Type 2 Diabetes and Cardiovascular disease treatment. Pipeline AOM therapeutic assets, more effective than Wecovy, are 1/15 the cost. Prepare for a paradigm shift.
Cedar Sinai does 1,000 valve replacements each year, which seems low.
Handheld ultrasound will replace stethoscopes, transforming healthcare. With AI, first-line diagnostic imaging will improve 100x. A tech with a handheld ultrasound will wipe out many undiagnosed diseases in less than five years.
Forget the Patriot Act; the feds are using satellites to find illegal manure spreading in Wisconsin (Article)
Prolon Fast Mimicking Diet is not that bad. 8 lbs in 5 days. I feel great.
Our food system is a gateway drug to healthcare. Farmers are giving up value to doctors. Farmers need to repackage their offerings in new business models focused on nutrition density. Take market share from healthcare.
Some Conclusions
Legal and General and the King's Fund in the UK (under the concept of ethical investment compliance) are trying to force food companies to reduce poor nutrition. While an appealing solution to affect the masses, people ultimately control their metabolic health, not CPGs. CPGs make money with better products, quality nutrition, better taste, and reduced metabolic harm. With little damage to sales even at 10% higher prices if consumers feel it reduces health costs.
Congressional Budget Office thinks current Anti-Obesity Medications (AOM) cost more than savings in Type 2 Diabetes and Cardiovascular disease. They, however, believe this will change in the near future.
Mini Mills
In the early 1980s, Japanese steel manufacturers put large US steel mills out of business, driving down costs despite long supply chains. In the late 1980s, Mini-Mills emerged in the US in part because they could more tightly time production line demand. The mini-mills were more local than Japan shipping routes, with less inventory and better quality feedback loops.
Will something similar happen in Agriculture?
Scale matters in commodity agriculture. Transportation and milling are significant costs for ingredients. The location of mills, rail/truck routes, and the location of demand matter. GMO/Non-GMO, protein density, and other attributes further confuse the supply chain flow. Grain can be stored, processed, and unprocessed. Supply comes in one big load in the fall. Offtake occurs throughout the year. Rather than Just in Time, it is more like Anything but in Time.
If CPGs start innovating nutrient-dense food, they also have to deal with demand forecasts. Will people like this new thing? How much? If the input is a custom SKU, it's hard to "Mix" from commodity. The supply chain then becomes a critical part of the cost/product decision. No one in CPG land has really had to worry about this because they mainly source commodity inputs.
90% of new products introduced by CPGs fail within one year of introduction. If Ozempic and other Anti-Obesity Medications are used more, experience suggests a 25% reduction in calorie demand and a 45% reduction in processed food demand. 90% of CPG product development talent is focused on value engineering, not nutrition optimization. Consumer pressure on climate has CPGs focused on farm carbon, not nutrition. If CPGs see a 45% reduction in demand in the future, worry that 90% of new products will fail, and have to optimize for climate, they have a lot to worry about all at once.
“How did you go bankrupt?” Bill asked.
“Two ways,” Mike said. “Gradually and then suddenly.”
Ernest Hemingway, The Sun Also Rises
It is a longer story, but the answer will be that CPGs need to concentrate on optimizing nutrient density. Nutrient density improves taste and cost. Its positive externality is carbon reduction. This means they need to source more novel soy, wheat, and/or corn. With more protein and/or micronutrients. In volumes of different sizes, in supply chains of different sorts. It means farmers need incentives to improve soil health becasue nutrient density is part plant genetics and part soil health. It means that CPGs need processors with the flexibility to mill to different demands that are a bit fickle. If you are a student of Lean and Six Sigma, continuous production flow, and cellular manufacturing, you have seen this episode before. And the answer is mini-mills. Big Steel never saw this coming, but firms like White River Soy do. And Edacious, a nutrient density assay company, provides the other critical technology needed to shift the supply chain to prioritize nutrient density.
I won’t complete the story today, but the “Tech Stack” to nutrient density and Food is Health includes:
Holganix - Soil Microbiome
Kula Bio - Nitrogen Fixation
Benson Hill - Genetics
White River Soy - Mini Mill
Edacious - Nutrient density optimization
But will it happen? You never know. So many ideas fall on the cutting room floor.
This reminds me that I still can’t get a good answer to whatever happened to that firm SpaceX from a 2004 article in Aviation Week. I recall us at Boeing talking about it, but we were focused elsewhere.
New in AI
Agent Hub
If you have been fooling around with AI and want to set up an RPA/IPA to process data and tasks, you should just go over to Agent Hub and get a subscription. It is early in their arc. They are brilliant. This is going to be a really good tool. In many ways what AutoGPT should have been. Their customer support and client engagement maximizes real-time feedback, which speeds product evolution. They have fully absorbed the YC lessons. I can slack them a question, and the programmers fix the code that night. My kind of entrepreneur.
This is Rahul coding up a fix for a customer while riding an Uber
I used their tool to build a trend-monitoring engine in less than an hour. It scrapes Reddit for top trending discussions in Farming, Health, and other subgroups. It pipes the top 100 posts from each subreddit from the last week into OpenAI or Opus. It looks at our investment criteria and list of startups. Then, it writes an analyst report on trending technology, customer unmet needs, and customer discussions about our startups. I linked it to a trigger 10 PM Friday night. It runs the whole thing while I sleep and sends me an email each week. The next step is to hook up LinkedIn with Phantom Buster and then Twitter/X. Probably $20 of computing to monitor 50 distinct data sources and update the team.
We built a system like this for Network-Centric warfare when I was at Boeing, and it only cost the government a couple of Billion.
On X…
Weeks Charts
I am optimistic. I graduated High School in 1985, when we had a mini-series about the US being destroyed by nuclear war.
This is one of the first computers I used to program. The picture is of Bill Gates and Paul Allen, but I did it in a similar room at Rye Country Day and then on a PDP/11 at Hotchkiss. And in a smaller classified SCIF at McDonnell Douglas. Bill and Paul stuck with it longer than I did, and I hear it worked out well for them.
These Harvard Students must be from Lake Wobegon. What happens when it hits 4.0?
So much good stuff here: "90% of new products introduced by CPGs fail within one year of introduction." - this must be a massive sword of Damocles for all product development.
Re Pessimism among 12th graders: Alarmism-for-money has reached its greatest form on TikTok. All of the lack of nuance of Twitter, and a picture paints a thousand words for 60 seconds.
Agent Hub team is awesome - trying to be like them.
Thanks for the tip about AgentHub—that looks great. You mention RPA here, which is interesting to me. It seems to me that AI agents will eat traditional RPA tools like UIPath, etc.